built a dynamic time warping demo and refactored the pure python dtw code into the demo

.gitignore

@@ -3,3 +3,4 @@
 *.log
 /build
 __pycache__
+.ipynb_checkpoints

dtw_pure_python.py

-import numpy as np
-
-def multi_dtw_pure_python(Xs, Ys, delta):
-	D = np.zeros((len(Xs), len(Ys)))
-	for i in range(len(Xs)):
-		for j in range(len(Ys)):
-			D[i, j] = dtw_pure_python(Xs[i], Ys[j], delta)
-	return D
-
-def dtw_pure_python(x, y, delta):
-	# First, compute all pairwise replacements
-	Delta = np.zeros((len(x), len(y)))
-	for i in range(len(x)):
-		for j in range(len(y)):
-			Delta[i, j] = delta(x[i], y[j])
-	# Then, compute the dynamic time warping
-	# distance
-	D = np.zeros((len(x), len(y)))
-	# initialize last entry
-	D[-1, -1] = Delta[-1, -1]
-	# initialize last column
-	for i in range(len(x)-2,-1,-1):
-		D[i,-1] = Delta[i,-1] + D[i+1,-1]
-	# initialize last row
-	for j in range(len(y)-2,-1,-1):
-		D[-1,j] = Delta[-1,j] + D[-1,j+1]
-
-	for i in range(len(x)-2,-1,-1):
-		for j in range(len(y)-2,-1,-1):
-			D[i,j] = Delta[i,j] + min3(D[i+1,j+1], D[i,j+1], D[i+1,j])
-	return D[0,0]
-
-def min3(a, b, c):
-	if(a < b):
-		if(a < c):
-			return a
-		else:
-			return c
-	else:
-		if(b < c):
-			return b
-		else:
-			return c

edist/multiprocess.py

@@ -68,10 +68,6 @@ def pairwise_distances(Xs, Ys, dist, delta = None, num_jobs = 8):
     # set up the result matrix
     D = np.zeros((K,L))
 
-    # set up the callback function
-    def callback(tpl):
-        D[tpl[0], tpl[1]] = tpl[2]
-
     # start off all parallel processing jobs
     results = []
     if(not delta):